Currently alignment techniques are known, e.g. in the field of semiconductor manufacture, where an alignment with a positional accuracy of about 100 nm can be achieved with optical techniques. However, it is very difficult to achieve an alignment accuracy significantly better than 10 nm because of the diffraction limit which applies to the resolution of optical alignment techniques.
There are some semiconductor manufacturing processes involving nano-structures which would benefit from an alignment accuracy of better than 10 nm and preferably of around 1 nm which cannot currently be achieved, or can only be achieved with significant technical complexity. One such application is the manufacture of semiconductor circuits using templates and curable resins to define the circuit patterns which are to be realized. Such manufacturing processes are summarized below.
There are also other technical fields in which a high alignment accuracy is either currently desirable or which could benefit from such high alignment accuracies, if suitable methods and apparatus for achieving alignment accuracies of tens of nm or better were available in a production environment. For example, there is increasing interest in nano-electronics, molecular electronics, single electron devices, or microfluidic devices which could be made significantly smaller if higher alignment accuracies could be achieved.
Similarly applications for high precision alignment apparatus and methods are conceivable in the biological or chemical fields. For example one could conceive a holder for biological or chemical use having an array of regularly (or irregularly) positioned recesses on a nanometer scale, each containing a sample or a reagent which has to be brought together with high positional accuracy with a carrier having reagents or samples positioned on a complementary array of projections which have to engage in the aforementioned recesses.
In nanoimprint lithography in which a negative three dimensional pattern provided on a template, for example of fused silica, is transferred to a thin layer of silicon containing monomer on a semiconductor or insulating substrate which is subsequently polymerized by UV illumination to form a hard positive pattern of the cured polymer on the surface of the substrate. In this technique the cured polymer is subsequently etched to remove a residual layer of polymer between raised features of the pattern and to reach the substrate material at these positions. Thereafter, the substrate may be etched further to produce depressions in the substrate material, and increase the aspect ratio of the raised features relative to the depressions, i.e. the depth of the depressions relative to the raised features. Then the residual polymer can be stripped from the substrate and one or more layers of semiconducting or insulating or conducting material deposited on the substrate. Following this, and appropriate polishing of the surface of the substrate, an organic planarization layer is deposited on the substrate and the process is then repeated using a different template and a new layer of UV-curable imprint solution. This process is then repeated for further templates, for example frequently using twenty or thirty different templates to produce the finished semiconductor circuit. This process, known e.g. under the name S-FIL™, a registered trade mark of Molecular Imprints Inc., is discussed, together with other lithographic processes, in more detail in an appendix to the present application.
The templates used are for example available to order from firms such as DuPont Photomasks and Photronics, who currently take orders for S-FIL templates down to 100 nm feature sizes. Only the template fabrication process, typically accomplished with an e-beam writer, limits the resolution of the features. Features as small as 20 nm have been made to date that exceed the present requirements specified in the International Technology Roadmap Semiconductors (ITRS). With this background in mind it will be appreciated that, although there is generally no critical alignment problem with applying the first imprint to a substrate using a first template, any subsequent template requires critical alignment with the pattern determined by the first and succeeding templates if the semiconductor circuit is to have any chance of operating as desired. The present invention provides such a tool.
In known processes of manufacturing semiconductors surface relief markings are regularly applied to semiconductor wafers to enable alignment of a series of masks or imprint templates with the wafer. They are however recognized and used for alignment by optical systems which, as explained above, have a diffraction limited resolution limit of about 100 nm. The same surface relief markings can be used for the purposes of the present invention. However, it can be preferable to make them smaller and for to position them with a smaller pitch. The surface relief markings can also be made with special topographies which enable them and/or their position to be detected more accurately and for reliably.
One proposal for achieving high accuracies in the alignment of a patterned mold (a first article) with a substrate (a second article) suitable for the manufacture of a semiconductor structure by an imprint process is described in the published international patent application WO 02/077716. That document describes a lithographic method which comprises aligning a patterned mold with respect to an alignment mark disposed on the substrate. The detection process is based upon interaction of a scanning probe with the alignment mark. The alignment mark can be formed by the edges of relief features provided on the substrate outside of the area to be patterned.
In the system described in WO 02/077716 an optical alignment system is first used to approximately align the patterned mold with the substrate and the precise alignment of the patterned mold with the substrate is then effected by a scanning probe alignment system either realized as a scanning tunnel microscope scanning assembly, in which the positions of the probes and thus of the patterned mold are controlled based on tunneling current information, or implemented as an atomic force microscope scanning assembly, in which the positions of probes are controlled based upon a force (e.g. an atomic force, an electrostatic force or a magnetic force) that is generated between the probes and one or more alignment marks carried on the substrate.
More specifically the scanning system is configured to move a scanning head, which carries the patterned mold and the probes, precisely in a plane, the x-y plane, that is parallel to the support surface of a stationary block carrying the substrate. The scanning system is also configured to move the scanning head precisely in the z-direction which is orthogonal to the support surface of the stationary block. It is stated that in one embodiment the scanning head may be moved vertically by a z-axis scan actuator and horizontally by a separate x-y axis scan actuator.
The actuators can be implemented as planar electrostatic actuators and can both be carried on the scanning head. After scanning the alignment mark in the x-y plane using the x-y scan actuator the precise position of the alignment mark relative to the patterned mold can be determined and the x-y actuators used to move the patterned mold into the desired alignment with the alignment marks and thus the substrate for the patterning process. The z-actuator can then be used to move the patterned mold vertically to impress the pattern thereon into a moldable film provided on the substrate. It is also stated that the probes can be retracted after alignment prior to the movement in the z-direction. However, it is not explained how this can be done.
The problem with the scanning system described in WO 02/077716 is that the detection of the alignment marks through scanning movement of the probes takes a relatively long time. This means that the manufacturing process for the semiconductor structure, which can involve the use of many imprint steps, with processing steps following each imprint step thus requiring repeated realignment, takes a relatively long time, which is undesirable. It should be appreciated that this applies irrespective of whether the scanning system is realized as a scanning tunnel microscope or is based on an atomic force microscope scanning assembly in which the positions of the probes are controlled based upon a force, such as an atomic force, an electrostatic force or a magnetic force. As those skilled in the art of atomic force microscopes will know these are all time-consuming non-contact measurement techniques. U.S. Pat. No. 5,317,141 describes a similar system to WO 02/077716. The principal difference is that the US patent is concerned with the alignment of a mask for X-ray lithography with a wafer, for subsequent patterning of the wafer using x-ray beams directed through the X-ray mask. Again a scanning probe microscope is used, e.g. in the form of an atomic force microscope which functions by scanning a fine-tipped probe over the surface of an alignment mark on the wafer or substrate. More specifically, a voltage difference is applied between the probe and the alignment mark and the tunneling current which results when the probe is a small distance from the surface is detected. For this system the alignment mark must have a conductive surface.
This is again a non-contact measurement. The detection of the tunneling current during scanning of the probe over the alignment mark is effected by a piezoelectric block carrying the probe. Control voltages can be applied to electrodes on the piezoelectric block to first energize it to move the probe in the z-direction to detect a tunneling current as the probe approaches the surface of the alignment mark. Thereafter further control voltages can be applied to appropriately positioned electrodes on the piezoelectric block to produce scanning movement of the probe in the x- and y-directions. The variation in the tunneling current and thus the topology of the alignment mark can then be determined during the scanning movements. This allows the position of the sensing head relative to the alignment mark to be determined with high accuracy. Because of the restriction involving the need for the alignment mark to have a conductive surface the US patent also discloses a second system in which a contact arm touches the surface of the alignment mark and the sensing tip is carried by the piezoelectric block at a small distance above the arm. The arm is electrically conductive so that the tunneling current can be measured between the arm and the tip of the probe which is spaced from the arm. The distance between the arm and the tip of the probe varies as the probe and arm are scanned over the surface of the alignment mark. Again the scanning of the alignment mark is relatively slow.
For the sake of completeness reference should also be made to two further documents which refer generally to atomic force microscopy. The first is DE-A-103 03 040 which describes a non-contact mode detector incorporated in a cantilever. This non-contact mode detector is also described, together with other non-contact detectors, in the paper: Micromachined atomic force microscopy sensor with integrated piezoresistive sensor and thermal bimorph actuator for high speed non-contact mode atomic force microscopy phase imaging in higher eigenmodes by R. Pedrak, Tzv. Ivanov, K. Ivanova, T Gotszalk, N. Abedinov, I. W. Rangelow, K. Edinger, E. Tomerov, T. Schenkel and P. Hudek in J. Vac. Sci. Technol. B 21(6) November/December 2003 pages 3102 to 3107. More specifically the above referenced article describes microprobes for non-contact scanning force microscopy, more specifically tapping mode atomic force microscopy. In this arrangement a cantilever carrying a tip is excited to oscillate close to its resonance. The topography information is collected from the phase lag between vibration excitation and response of the cantilever deflection sensor.
In one embodiment described in the above referenced article the cantilever is realized as a bimorph actuator involving an aluminium layer on a silicon dioxide cantilever. The aluminium layer can be heated with an oscillating current to produce oscillatory bending deflection of the cantilever due to differential thermal expansion. In one embodiment described in the paper the cantilever is configured to include a piezoresistive detector realized in the form of a Wheatstone bridge.
The production of probes for atomic force microscopy, including a probe which utilizes a piezoresistive Wheatstone bridge is also described in the document SPIE Vol 2879/pages 56 to 64, being a paper presented at a conference in Texas on Oct. 14-15, 1996. A piezoresistive detector incorporated in a cantilever is also described in U.S. Pat. No. 5,444,244.